How to Use an AI Assistant with Your Monitoring System

As observability systems grow more complex, so does the cognitive load on teams. This talk explores how AI assistants can act as intelligent interfaces to your monitoring stack using MCP (Model Context Protocol). You’ll see how teams can query metrics, logs, and traces in plain English, debug issues faster, analyze alerts and cardinality, and work with configurations using natural language. We’ll cover MCP architecture, practical setup, real-world use cases, and lessons learned from integrating AI into daily workflows. This approach doesn’t replace expertise, it amplifies it, making advanced observability accessible to entire teams.